UC San Diego

The ability to sample random networks that can accurately represent real social contact networks is essential to the study of viral epidemics. The Barab asi-Albert model and its extensions attempt to capture reality by generating networks with power-law degree distributions, but properties of the resulting distributions (e.g. minimum, average, and maximum degree) are often unrealistic of the social contacts the models attempt to capture. I propose a novel extension of the Barab asi-Albert model, which I call the "dual-Barab asi-Albert" (DBA) model, that attempts to better capture these properties of real networks of social contact.